Designing with Arc-Flash in Mind

[JoeStillman]

The local POCO publishes available fault currents for typical transformer secondaries so we can do a SC calc up-front and specify AIC ratings. This is very helpful because the primary infrastructure may not even be designed yet. But what about the Arc-Flash?

Imagine a 600A 480 V service to a new building. My model starts at the transformer secondary and has unprotected cable up to the service main. Using the published max SC current at the transformer of 40 kA and 100' of cable, my SC at the service entrance is 30 kA and I can use 35 kAIC gear. But my arc-flash will be over 100 calories, which means an unsafe-for-live-work sticker. In that scenario, I would want to use a main fused switch feeding a MLO panel instead of a main breaker panel so that work in the future can happen inside the panel.

But if I include in my model an assumed infinite utility feeding a 500 kVA transformer with a 1.5%Z and a C12 Bay-O-Net fuse, (never-mind any back-up fuse), I'm at 5.4 calories on the line-side of my main disconnect and my sticker goes from red to orange. What should I do? Assume the worst and include possibly redundant equipment in the design? Or assume the utility will include the protection I imagine on the primary? That's too many "assumes" and no "knows".

Does anybody else make assumptions about the utility primary protection, before the utility company has designed anything?

 

[petersonra]

Have you considered asking the POCO what they plan to supply?

 

[jim dungar]

The local POCO publishes available fault currents for typical transformer secondaries so we can do a SC calc up-front and specify AIC ratings. This is very helpful because the primary infrastructure may not even be designed yet. But what about the Arc-Flash?

Does anybody else make assumptions about the utility primary protection, before the utility company has designed anything?

Other than running several different scenarios, like smaller transformer and lower avaliable fault current, you don't have much choice for the line side of your main device.

For many years now, the local marketplace has simply used a remote main device, typically a breaker, feeding a main lug panel. The service entrance device may still have a high incident energy but the panel is usually below 8 cal/cm², regardless of the final utility values.

 

[JoeStillman]

Have you considered asking the POCO what they plan to supply?

Yes, but there are so many gatekeepers between the general public and the POCO engineers that it can take weeks just to find out who to even ask. It's almost impossible to get someone on the phone who even understands the question.

 

[jim dungar]

Have you considered asking the POCO what they plan to supply?

The six major companies, in this area, will not provide any fault current data, for use with arc flash calculations, to customers purchasing power at 480V or less.

 

[paulengr]

The problem is AIC is almost easy. You design for the highest SCCR and that’s it.

But with arc flash as current goes down, incident energy goes up. So unlike AIC you can’t predict it ahead of time at least on the line side of the main. But knowing the current at 2 seconds on the breaker curve, incident energy reaches a maximum and we can predict downstream.

 

[Fred B]

The local POCO publishes available fault currents for typical transformer secondaries so we can do a SC calc up-front and specify AIC ratings. This is very helpful because the primary infrastructure may not even be designed yet. But what about the Arc-Flash?

.................. What should I do? Assume the worst and include possibly redundant equipment in the design? Or assume the utility will include the protection I imagine on the primary? That's too many "assumes" and no "knows".

Does anybody else make assumptions about the utility primary protection, before the utility company has designed anything?

I try to never assume anyone, especially the POCO, would do anything. If you are in planning stage, I plan worst case scenario, not sure of price point difference between getting your Arc flash from red to orange would be, but if you install "under designed" and POCO doesn't design to your expectations, life/safety are potentially at risk.
Even after design and installation, I've seen the POCO make alterations later on that would potentially effect the AIC and they dont inform anyone down stream that changes are taking place. I'm not sure if they were to make a drastic change that effects significantly the Arc flash potential that they would make contact with down stream individuals.

 

[wbdvt]

First off, what the POCO published was not available fault currents. What you received is infinite bus fault currents which are appropriate to use for design to determine the needed SCCR and AIC ratings of the equipment. You will need the available fault current to calculate the incident energy and the POCO should be able to supply that to you since the location of the service would probably be known if you are designing the electrical system. Some assumptions can be made such as length of riser cable to padmount txf, riser fuses would be in accordance with POCO design standards, secondary run can be estimated.

I remember seeing an IEEE paper, I think Dow Chemical engineers published it, and the worst case arc flash hazards were when the fault current was 50% of the infinite bus current.

[ The six major companies, in this area, will not provide any fault current data, for use with arc flash calculations, to customers purchasing power at 480V or less. /QUOTE]

I ran this before with a utility and ended up filing a complaint with the state PUC about the utility not providing the necessary information to allow my client to meet OSHA requirements to protect their employees. IEEE 1584 was used to show that the available fault current is needed to accurately determine the magnitude of the hazard so that properly rated PPE could be used. Examples were provided that showed that use of the infinite bus fault current as provided by the utility actually resulted in incident energy values that were less than actually there which would have left an employee seriously injured or killed.

 

[paulengr]

First off, what the POCO published was not available fault currents. What you received is infinite bus fault currents which are appropriate to use for design to determine the needed SCCR and AIC ratings of the equipment. You will need the available fault current to calculate the incident energy and the POCO should be able to supply that to you since the location of the service would probably be known if you are designing the electrical system. Some assumptions can be made such as length of riser cable to padmount txf, riser fuses would be in accordance with POCO design standards, secondary run can be estimated.

I remember seeing an IEEE paper, I think Dow Chemical engineers published it, and the worst case arc flash hazards were when the fault current was 50% of the infinite bus current.

For typical breaker curves (everything except moderate inverse) if you are on the inverse time curve as current decreases the time increases so much that the incident energy increases.

If you are in the instantaneous region or at a point where trip time exceeds two seconds, time is fixed so incident energy does what you expect.

 

[Bwas]

In general I make no utility assumptions for preliminary calcs. If you guess wrong or the utility does something unexpected it can change everything and often times there is nothing you can do.

I had one industrial client where I measured overhead line distances back to the substation and included fusing at the transformer, but it was a transformer I specified. The client had a "Maximum allowable arc flash of category 2" and the electrical manager needed to be able to tell his management we did everything possible to get down to that level, then they granted an exception. I was working with an informed client and we were on the same page on what the results would probably be.